Treatment of Non-APL Acute Myelogenous Leukemia with Intravenous Arsenic Trioxide Plus Ascorbic Acid.

Abstract Introduction: Arsenic trioxide (ATO) has demonstrated efficacy in acute promyelocytic leukemia (APL), inducing complete remissions (CR) in 85% of relapsed patients. On the other hand, ATO alone has no activity in non-APL AML. Although this difference could be explained by degradation of the APL specific PML/RAR fusion protein, ATO has shown to be active against APL cells by mechanisms that are independent of this gene rearrangement. ATO-induced apoptosis correlates inversely with intracellular concentrations of reduced glutathione (GSH) via generation of reactive oxygen species (ie, cells with high concentrations of GSH are more resistant to ATO). APL cells were found to have low levels of GSH compared to non-AML cells. Agents that reduce intracellular GSH, such as ascorbic acid (AA), increase ATO-induced apoptosis and can overcome resistance to ATO in non-APL AML cell lines; AA alone has no activity in these cells. Thus, the combination of ATO and AA was evaluated in patients with non-APL AML. Methods: ATO (0.25 mg/kg/d) was administrated intravenously over 1–4 hours with intravenous AA (1g/d, given 30 minutes after ATO) for five days/week (five days on/2 days off) for five weeks (25 doses=one cycle). These doses were based on a phase I/II trial of ATO + AA in patients with multiple myeloma (Behalis et al Clin Cancer Res8:3658, 2002). Responding patients received an additional consolidation cycle of 25 doses followed by maintenance including ATO + AA two weeks of every month for 4 cycles. Patients who failed to respond after two cycles were considered treatment failures. Results: Eleven patients, aged 36–84 years (median: 66 years), were enrolled: 5 had relapsed after chemotherapy; and 6 elderly, aged 66–84, (including 5 with antecedent MDS), were previously untreated. One patient was discontinued prematurely because of QT prolongation. Major response was seen in 3 (27%) pts: CR or CRp in 2 patients and decrease in bone marrow blasts to <5% with reduced transfusion requirements in 1 patient. Three additional patients had a drop in peripheral blood blasts from 85% and 51% to <5%. All responding patients, except for 1, were previously untreated. Responses were observed after the first treatment cycle. Duration of responses were 6 months in the CR and CRp patients, and 1 week, 2 months and 4 months in the other responders. Four pts failed and had progressive disease; one patient died on study from progressive disease and infection. Grade 3/4 toxicity included infection (n=8), and QT (>500 msec), anorexia, fatigue, sensory neuropathy, elevated bilirubin (n=1 each). One responding patient developed shortness of breath with severe hypoxemia, reminiscent of the APL differentiation syndrome, which responded immediately to dexamethasone. Conclusion: ATO + AA can have anti-leukemic activity in poor risk non-APL AML with much less toxicity than chemotherapy, especially in previously untreated patients. Most relapsed patients did not respond to ATO +AA, suggesting that chemotherapy can induce resistance to this combination. Further studies are needed to explore the potential of ATO +AA as a front-line therapy instead of intensive chemotherapy in elderly patients with non-APL AML.